Design leadership, AI-native product, and what changes when agents join the team.
People ask me how I actually use AI tools in my design process. I split the work three ways. The first 10% is mine. The middle 70% is where AI does the heavy lifting while I steer. The last 20% is mine again.
I mapped every live screen in our product, grouped them into flows, and connected analytics. This is what it showed me about entry points and disclosure patterns, and why the map is the layer AI should read first.
Craft did not disappear. It moved to the judgments that decide what deserves to ship.
The system’s new home is Storybook. Figma is upstream. Most startups should not build a system from scratch. They should start from Meta’s Astryx.
The biggest blocker to AI adoption on design teams isn’t the technology. It’s overwhelm, and underneath the overwhelm, something quieter: fear. This is how we designed the culture, and the hackathon, that closed the gap.
Most of a design leader’s week is invisible coordination work. Claude Cowork is the first AI tool that can credibly take a real slice of it. This is how I am using it, what it is genuinely good at, and where it still falls short.
Most AI product writing is hypothetical. These principles come out of shipping AI to tens of millions of users at Credit Karma. Use them, argue with them, but do not skip them.
There is a new AI design tool every Tuesday. Most do not survive contact with a real team. This is the stack that stuck, the tools we dropped, and the rule we use to judge the next one.
Most hackathons end with a slideshow and zero shipped work. This is the format I use to reskill design teams on AI and walk out with prototypes the product team actually wants.
NPS and CSAT will not get you promoted, and they will not save your team in a cut cycle. Here is the small set of metrics I use to make design measurable to the people who control headcount.
Most founders hire a Head of Design too late, then optimize for the wrong signal. This is what to look for, how to interview for it, and the first 90 days that will tell you whether you got it right.
Pods, rituals, and roles built for waterfall hand-offs are quietly breaking under AI-native product work. This is the structure I am moving teams toward, and the order of operations to get there without burning the org.
Most designers misread the ladder. The jump is rarely about better Figma files. It is about what you are trusted to decide, who else’s career you are accountable for, and how your calendar fills up.
AI tools and 10x goals are creating real anxiety for designers, employed or not. The fix is not another all-hands speech. It is hands-on reskilling, hackathons with engineers, and making learning a visible part of the work.
When every function optimizes for its own scoreboard, the product loses. This is how the best teams align incentives, and what they measure together.
A four-phase maturity roadmap I use, and teach design leaders to use, when growing a team from a request queue into a product function. Each phase has its signals, its moves, and the work it unlocks.
Most startups start with aesthetics. Product success is built from the bottom up. Seven layers of product experience, prioritized by what actually drives adoption, retention, and revenue.
A five-step framework refined over 11 years of leading design in fintech, marketplace, and AI-first products. Each step compresses ambiguity into clarity, and clarity into measurable outcomes.
Operating principles refined over 12+ years of leading design in fintech, marketplace, and AI-first products at Intuit, Credit Karma, and LendingClub.
A working framework I use, and teach my teams to use, when building products in an AI-native loop. Five questions, in order, from validating the problem to trusting the funnel.
I write about design leadership, AI-native product craft, and the messy reality of shipping at scale. Get notified when something new drops.